Michael J. Mauboussin of Legg Mason Capital Management has an excellent article (pdf) about common errors in DCF models. Below is a summary of the eight main points and couple of quotes from the article (pdf).
- Forecast horizon that is too short.
- Uneconomic continuing value.
- Cost of capital.
- Mismatch between assumed investment and earnings growth.
- Improper reflection of other liabilities.
- Discount to private market value.
- Double counting.
- Scenarios. An intelligent investor needs to consider multiple scenarios. Investors should look to the value drivers—sales, margins, and investment needs—as sources of variant perception. Proper scenario analysis considers how changes in sales, costs, and investments lead to varying value driver outcomes.
Quote about valuation multiples
Multiples are not valuation; they represent shorthand for the valuation process. Like most forms of shorthand, multiples come with blind spots and biases that few investors take the time and care to understand.
General conclusion
Theory and practice tell us the value of a company is the present value of future cash flows. Investors primarily seek to buy a stream of cash flow for less than it’s worth—or sell a stream for more than it’s worth. Accordingly, an investor needs to be able to model cash flows intelligently and identify a variant perception: a well-founded belief the market has placed an incorrect value on a company.
bert frenso 
bertfresno, I think I agree with all of Michael J. Mauboussin’s points he made about common errors in DCF models. I have seen in the economic/finance literatures more advanced methods to DCF models which are more robust (minimal errors) than the bare DCF itself. Some of the more advanced valuation methods are basically an extension of the classical DCF. I’ve already mentioned some (there are many available today) of those techniques on Jim Donovan’s blog.