Nonlinear regression works iteratively. Prism starts with initial estimated values for each parameter. It then gradually adjusts these until it converges on the best fit. In rare circumstances, Prism cannot compute a curve based on the initial values you entered. When it tries, it encounters a math problem such as division by zero or taking the logarithm of a negative number. In this case, Prism does not report any results and instead reports "bad initial values".
Did you enter or choose the right equation?
Did you enter sensible values for initial values?
Is the range of X values invalid? Focus on the first and last X value (and on X=0 if included in the range). Can the equation be evaluated at those X values with the initial values you entered?