No matter what your present level of leadership, the first hundred days of a new job holds both a unique window of opportunity and a heightened state of risk. The opportunity comes from the fact that a leadership transition is a time of maximum uncertainty, when all assumptions are up in the air and open to change. As the new leader, you’re generally given the benefit of the doubt. “I think it’s you’re one chance to ask stupid questions,” says Robert Eckert, chief executive of Mattel, who had an exceptional first hundred days.

And while some of the specific do’s and don'ts for CEOs are unique to their role, most essential things like setting expectations, developing a vision, establishing a management process, creating priorities, building your management team, and doing an exceptional job on your earliest projects apply equally to anyone in a new leadership role.

When Ron Daniel was the managing partner of McKinsey & Company from 1976 to 1988, he sent a memo to new recruits when they started, entitled, “On Becoming an Associate.” His advice is still memorable all these years later: “Recognize the necessity of getting off to a good start in the firm. Your first few engagements are critical. During these studies, you can establish an internal clientele for yourself - that is, by performing in an outstanding way, your reputation will be quickly established in your office and even the firm.” (We’ve incorporated Ron’s memo in the Appendix of the book.)

This is especially important because whenever you assume a new role, you’re in what Max DePree, former CEO of furniture company Herman Miller and author of Leadership Is an Art, calls “a temporary state of incompetence.” Even if you think you know a company - or a department or a division - before you take over its leadership, think again. As GE’s Immelt reminisces, “I worked for this place for twenty-one years before I got the CEO job and there were still things that shocked me when I took over.”

The knowledge gap is even wider for outsiders. “Anyone coming into a new situation is faced with the fact that they often have to do the most at a point they know the least. You may have previous experience and you may be smart and have insight into how things work, but you know the least about the actual company you’re engaged in at the same time you have to set things in motion,” says AOL chairman and CEO Jonathan F. Miller, himself recruited into the company from the outside.

An identity crisis knocks a company’s compass off kilter. It clouds its people’s ability to make clear decisions, to choose which route to take, and to allocate the proper resources. How can you hire the right people if you don’t know what you’re hiring them to do? How do you know which projects and products to support if you don’t know how they will ultimately fit into the whole?

Almost every leader will say that their lodestar is defining their business focus, which comes down to what you can do to build long-term value. “If anything, what the last few years have taught us is that building long-term value is critical for any company,” says Dave Peterschmidt. “If the company is going to be in there for the long haul, it has to understand the core value it brings to the market and the reason for its existence in a crowded marketplace. Then you can do an assessment to get the company tracking toward long-term values.”

Think of that assessment as conducting a reality check of the operating environment. If you want to get the company tracking toward long-term value, consider what conditions pertain that will help or hinder you. The answers will shape your short-term agenda.

In fact, Lou Gerstner claims that his greatest ally in breaking free from the past was IBM’s own precipitous collapse. But he also knew that the task of shaking people out of their stupor and getting them “to think and act collaboratively, as hungry, curious self-starters,” would take at least five years - and even then, he confesses, he underestimated the time required.

At the other end of the spectrum, if you have an organization that’s generally working pretty well, you want to protect the part that’s working by going a little slow, as McNerney did.

Lucent’s Schacht offers sound advice for the new CEO in a crisis:

“If I were a new CEO coming into a new company at a tough time. I’d have a deal with the board. I would tell them, ‘I’m going to keep you informed every step of the way. But you’re not my most important problem right now. You’re asked me to come in and you’ve given me a good briefing and I thank you; now please give me a little room. I’ll be back with a status report in two or three weeks, and if something else comes up, we’ll be on the phone together. But right now my most important constituent is the internal folks and our customers.