Presents the methods of computing full-information maximum likelihood (FIML) estimates in simultaneous stochastic equations. Derivation of normal equations for FIML estimates; Treatment of identities and linear restrictions; Presentation of numerical example.