Objective To discuss the relevant factors affecting the prognosis of cryptococcal infections of central nervous systeminnon-acquired immune deficiency syndrome(AIDS)patients and to establish effective and reasonable forecast models. Methods With retrospective analysis of 100 cases with cryptococcal infections of the central nervous system in non-AIDS patients, multivariate logical regression analysis was used to conduct statistical analyses of the data. Results The underlying diseases(x₁₀), parenchymal involvement(x₄), the glucose ratio of cerebrospinal fluid(CSF)to the blood(x₇), CSF cryptococcus cultivate positive rate(x₁₃), and the cryptococcus ink staining smear positive rate(x₁₂)on the initially collected CSF, were closely correlated with the outcome(y). And thelinear regression equation was: y=1/[1+exp(-z)], of which z=0.781x₁₀+2.287x₄+0.707x₇+1.134x₁₃-1.837x₁₂-4.420. By means of the hybrid tests, e.g. the Hosmer-Lemeshow test, it was found that the model was reasonable and that the average accuracy was as high as 84% based on sample tests. Conclusion This study shows that the prognosis of the cryptococcal infections of the central nervous system innon-AIDS patients is largely correlated with the host underlying conditions, parenchymal involvement, cerebrospinal fluid glucose drops, positive rate of cryptococcus culture, and ink staining smear on the initially collected CSF.