Working Through Kreps with a Solutions Manual
The Kreps textbook is standard for graduate micro theory. The solutions manual that circulates covers exercises from all eight chapters, though the quality of the solutions themselves varies depending on which version you find. Most students don't realize there are different editions floating around. The 1990 edition has a different exercise numbering than later printings, so what you see online might not line up with the book on your desk. I spent two semesters grading problem sets built off this material, and I can tell you the most common failure point isn't understanding the math. It's knowing which theorem applies when. Students will sit there staring at a sequential game problem, trying to force backward induction onto something that actually needs a different equilibrium concept. The solutions manual walks through these cleanly, but only if you know how to use it without cheating yourself out of the learning process.
Using the In Microeconomic Theory Kreps Solutions Manual Effectively
Here's what actually works. Open the problem first. Try three or four minutes on it without looking at anything. If you're stuck after that, peek at just the setup in the solution—the part before the algebra kicks in. This tells you whether you were approaching it from the right direction or completely lost. Then close it and work through the derivation yourself. Reading a solution and nodding along gives you the illusion of competence. Writing it out separately is where actual understanding happens. Chapter 2 on constrained optimization trips people up more than any other section. The Kuhn-Tucker conditions are presented tersely in Kreps, and students rarely internalize the complementary slackness condition until they've seen it applied in three or four different contexts. The manual handles this well in exercises 2.4 and 2.6, but I've noticed that some online versions skip the geometric interpretation that the original printed solutions include. Without that picture, you're memorizing algebra instead of understanding what the conditions mean. I ran into a specific issue last year with exercise 4.11 in the chapter on decision theory. The solution provided used a particular normalization of the utility function that wasn't stated explicitly, and graders who weren't paying attention marked students wrong for deriving the same result through a different normalization path. The workaround was simple: document each normalization choice on your page and state the equivalence at the end. Professors who care about correctness over notation will accept both approaches.
The dynamic programming section in Chapter 7 is where the manual really earns its keep. Kreps covers the contraction mapping approach briefly, and the exercises assume familiarity that most students don't have yet. Working through the Bellman equation derivations in the solutions alongside your own attempts for about an hour per exercise will cement the material faster than rereading the chapter. I've seen students cut this down to twenty minutes by being more selective—focus on exercises where the solution method differs from what you initially tried. There are real limitations to relying on this manual. The solutions sometimes skip intermediate steps, particularly in the general equilibrium chapters. When the manual jumps from one equation to another without showing the substitution, it's usually because the author assumes you've already verified the algebra. If you catch yourself confused, go back and fill in the steps rather than accepting the gap. Another problem: the later editions corrected some errors from the first printing, so if you're using a solutions manual from 1990 alongside a newer textbook version, you may encounter contradictions. Cross-reference the errata list that Kreps maintains before assuming the manual is wrong. For the exercises on repeated games in Chapter 11, the manual tends to present the folk theorem proof in a compact form that works for infinite horizon cases but doesn't address the finite horizon variant that shows up on exams. I've seen this cost students points because they applied the infinite-horizon result to a finitely repeated game without noting the distinction. The fix is straightforward—add a sentence or two acknowledging the difference and explaining why the one-shot deviation principle doesn't yield the same results when the game ends at a known date.
Get the Full Details

Download links for this manual circulate on academic forums and file-sharing sites, but the versions vary significantly. The most reliable source I've found is the Stanford course pages that linked to the official supplemental materials when Kreps was teaching there. Anything else requires checking that your exercises match your textbook edition before you invest time working through them.