Non-consequentialist theories claim that what we ought to do is not (or at least not solely) determined by the value of the consequences of what we do.  There are different and conflicting  types of non-consequentialist theories.  Here we shall look at Kant’s deontology and at Divine Command Theory
 

Kant’s Deontology

One way to approach Kant’s theory, is to contrast it with Utilitarianism.  For the latter, moral value and moral agency are derivative from non-moral value, i.e., the good: what's right and what one ought to do depend on how much pleasure is produced.
By contrast, for Kant moral value and moral agency are not derivative from non-moral value: morality is, in this respect, autonomous.
In addition, all rational beings are always and necessarily  duty bound by the moral law: that is, the moral law is universal and necessary.   From this, Kant infers that:

• Since one is morally obliged to do X only if it is in one's power to do X (in other words, for K. 'I ought' implies 'I can'), all rational beings can always satisfy that duty only if morality is internal to the agent in the sense that a good will (the intention and the sincere effort to obey the moral law) is all that's required.
• The ground of the moral law cannot rest on any consideration of what's empirically true of man (or any other rational being), but on a-priori (i.e., non-empirical) considerations about what it is to have a rational will.

1) A good will is the only thing good without qualification independently of its  success. While other capacities, e.g., intelligence, gifts of fortune, talent, can  operate against morality (think of a clever thief), a good will always acts in  accordance with the principles of morality.
 NOTE:

• The value of a good will is independent of anything external to it: Kant is not  saying that the principles of good willing depend on a non-moral good at which   they aim; rather he is going to determine what a good will is by  appealing to the   principles of morality.
•  To will is not to desire, but to decide on a course of action; hence, a good will   can be identified with practical reason.
•  A decision is morally good only if taken for the sake of one's duty   towards the moral law, because the agent thinks that the intended action is   right.
•  Acting for the sake of duty is not merely acting in accordance with duty.  For  example, not stealing out of fear of punishment is not acting for the sake of   duty (out of respect for the moral law “Don't steal.”)  The same holds in the  case of helping someone out of love and not out of respect for the    moral law “Help other persons.”

Rationale:
morality is binding on all rational beings all the times; hence, it must be possible to be moral all the times, so that even the misanthrope can act morally towards others. This doesn't entail that we cannot do our duty towards humanity while loving  it (as long as our love is not the motive of our action), but that we can do our duty towards  it even while hating it.

• Only persons (beings capable of acting in accordance with maxims, i.e., general principles) have a will (practical reason) and can be moral or immoral; non-  persons, e.g., animals, are a-moral, i.e., they're outside the realm of morality (a cat is not evil for "playing" with the mouse).

1.  “Act only on the maxim through which you can at the same time will that it be a   universal law” (Universal law formula)
2.  “Treat humanity in your own person or in the person of any other never simply as   a means but always at the same time as an end.” (End in itself formula)

 NOTE: Although strictly speaking there is only one Categorical Imperative, one can   call any maxim which satisfies (1) and/or (2), e.g., “Don’t lie,” a categorical   imperative.

•  Hence, a hypothetical imperative carries no implication that one ought to do what it commands (I might give up the idea of having good business).  By contrast, a categorical imperative impliesthat one ought to do what it states.
•  Since any consequentialist theory recommends an action or a maxim (rule) on  the basis of the good it produces, it is based not on categorical  imperatives, but on hypothetical ones, and therefore must be rejected.

•  Rationale for the law:

it determines maxims which could be adopted by a group of rational agents  without assuming anything specific about their desires, capacities or social relations, which are merely contingent, dependent on fortune.

•  Application of CI

 Presumably, CI provides us with a test to determine what to will, that is, whether an action is moral, immoral or merely permissible.  However, how this  is supposed to work is far from clear, and Kant’s own examples don't help too much.

1. give an adequate formulation of the universal maxim M embodied in your action.  Let N be M's contrary (not M’s contradictory).

       example:
       M = every time it's to one's advantage to do so, one lies.
       N = at no time when it's to one's advantage to do so does one lie.
    NOTE: not “not at all times when it's to one's advantage to do so, one lies”, which does not mean the same as N.
2. Suppose everyone followed M; then suppose that everyone followed N.
3. If only one of M and N is pragmatically self-defeating, then that one is an immoral maxim and its contrary is obligatory.

Look at the example: M would be pragmatically self defeating if everybody  followed it (who would believe you?).  Hence, M is immoral and N is morally  obligatory, in fact, a perfect duty.
 NOTE: a perfect duty is one which holds for all agents in all actions with all possible other persons.  They are associated with rights (people have a right not be lied to, but no right to be helped).

B. Test for imperfect duties.
  Same as above with one difference: at (3) substitute "incompatible with some basic features of rationality" for "pragmatically self-defeating."
Example 1:

• M = neither help people in need, nor seek help when in need.
• N = help people in need and seek help when in need.

Example 2:

• M= always seek pleasure neglecting your talents and skills
• N= never seek pleasure neglecting your talents and skills

6) The "end in itself" formulation of CI says:  “Treat humanity in your own person or in  the person of any other never simply as a means but always at the same time as an  end.”
 Rationale:
 Since humans act morally and have a rational will, they have intrinsic objective  value and hence they are ends in themselves.  They are, one might say, the only  intrinsic moral values (happiness or pleasure are non-moral values)  Since the value  of everything else is parasitic on that of persons and a means is less valuable than  its end, treating persons as means amounts to denying that they are ends in  themselves, that is, persons.
The meaning of the maxim is made clearer by the following:

• Treating X as a means consists involving X in a course of action to which  X  would not in principle consent, that is, a course of action which is contrary to X’s maxims.

  Examples: promise breaking, lying, coercing, etc.
  NOTE: these are perfect duties
  Problem: how does one deal with cases of false consciousness?
•  Treating X as an end consists in supporting (at a minimum not impeding) X’s capacities to achieve his ends, as long as they don't involve treating someone merely as a means

  NOTE:
  These are imperfect duties.
  A rational will manifests itself in the link between one's values and what one does. Treating people as means weakens the link; treating them as ends  enhances it.
Problem: someone is convicted of stealing, and the judge proceeds to sentence her. Is it necessary that the judge be advancing the prisoner's ends in sentencing. But the prisoner protests that she doesn't want to be sentenced;  "if you sentence me-she says you will simply be using me for your (and society's) ends."  So, is Kant's theory even compatible with punishment?
Reply: The prisoner is treated morally as long as the principle on which the judge acts in sentencing is one that it would be reasonable for the prisoner to accept. So, even though the prisoner does not want herself to be sentenced, she shouldn't be opposed to there being laws against stealing; after all, thieves too don't want their property to be stolen. So she would will the principle, even if she is not very happy about its application in this case.

•  a rational being, to the extent it is really rational, will find out what the CI   is and follow it.

NOTE: however, humans are very prone to immoral behavior because of:
1. impurity, due to the fact that often we follow CI not strictly out of duty but form mixed (impure) motives e.g., self-interest, love, etc.
2. wickedness, due to the fact that often we subordinate moral maxims (i.e. CI) to non-moral ones
•  In following CI, a rational being is autonomous, since one follows the rational code of behavior which one, insofar as rational, has oneself created.

NOTE: This is analogous to following rational laws in one's reasoning.

9) Kant’s view culminates in what he calls “the kingdom of ends.”  Consider a  “systematic union” of all rational beings acting on the basis of good will, i.e., on the  basis of universal and morally necessary laws.  Then, since laws determine ends  with regard to their “universal validity, we shall be able--if we abstract from the  personal differences between rational beings, and also from the content of their  private ends-- to conceive...a kingdom of ends which is possible in accordance with  the above principles [i.e., CI].”
 John Rawls has proposed that we take Kant to ask which maxims we would want to be universal law not from the standpoint of our individual situation (talents, race,  sex, wealth, etc., including conceptions of non-moral good), but from that of a  rational person who doesn't know his own individual situation (this is another way  of saying that in justifying universal maxims, we cannot appeal to our, or anybody's,  individual situation).  Form behind this “veil of ignorance,” we all would choose  those maxims which guarantee that our dignity as persons be satisfied and we be  able to pursue our individual conceptions of non-moral good, whatever they might  be, thus combining the two versions of CI we have considered.
 

10) Problems
 

1.  Rigorism.

 Kant’s ethics seems case-insensitive .  For example, Kant claims that one  should never lie.  But what should one do when someone's life is at stake?    Should the Danish fisherman smuggling Jews to Sweden tell the truth to the Nazi  officer on the patrol boat?   Most of us would say “no”, but Kant says “yes”.   (Notice that being silent won't do here).  Kant might argue that telling the truth  is a perfect duty (the Nazi has a right not to be lied to) while helping others is an imperfect duty (the Jew hasn't a right to be helped).  But many would find this answer appalling.  Of course one should lie to the Nazi.
 Reply:
 Perhaps Kant overstated the requirements of CI.  “Lie when and only when it is   necessary to save the life of an innocent person” seems to pass the universal   law test because its range is small enough not to destroy the institution of promise keeping, without which the maxim would become self-defeating.
 Even so, it's hard to see how the maxim would satisfy the end in itself version of  CI.
 NOTE: this raises the further problem of the equivalence of the various versions of CI.
 
2.  ii. Abstraction.

 Kant’s CI is too abstract because it's unable to guide action, i.e., to tell us what   to do in many specific cases.  In this it compares unfavorably with    Utilitarianism.
  Reply:
  It is true that CI doesn't provide any automatic procedure to guide action, but to   a large extent this is true of Utilitarianism as well (the hedonistic calculus is   often prohibitively hard).  Moreover, Kant never denied that moral action requires   deliberation and good judgment.