Sampling precision was investigated for Tylenchulus semipenetrans juveniles and males in soil and females from roots and for citrus fibrous root mass density. For the case of two composite samples of 15 cores each, counts of juvenile and male nematodes were estimated to be within 40% of mu, at P 1,500 nematodes/100 cm(3) soil. A similar level of alpha was estimated for measurements of fibrous root mass density, but at a precision level of 25% of mu. Densities of female nematodes were estimated with less precision than juveniles and males. Precision estimates from a general sample plan derived from Taylor's Power Law were in good agreement with estimates from individual orchards. Two aspects involved in deriving sampling plans for management advisory purposes were investigated. A minimum of five to six preliminary samples were required to appreciably reduce bias toward underestimation of sigma. The use of a Student's t value rather than a standard normal deviate in formulae to estimate sample size increased the estimates by an average of three units. Cases in which the use of z rather than Student's t is appropriate for these formulae are discussed.