Two kinds of length-based Beverton and Holt spawning stock biomass per-recruit (SSB/R)-models are presented. The first allows the computation of the absolute value of SSB/R. The second estimates the index of SSB/R, also called ‘relative SSB/R’. Both models are extensions of the original age-based Beverton and Holt SSB/R-models. The SSB/R and its index can be assessed with reference to variations in the exploitation characteristics. These models were applied to the Lates stappersii (Boulenger) stock in Lake Tanganyika using fishing mortality as the independent variable. They produce an absolute value of SSB/R of 12.95 g, representing 22.4% of that of the ‘virgin’ stock for the 1993–1994 equilibrium-fishing regime. This level is indicative of an unsafe state of the stock in question. Moreover, the SSB for L. stappersii seems to be fragile under increased fishing pressure as, for example, doubling or tripling the current fishing regime would reduce the current SSB/R by 57.3 and 77.4%, respectively. In general, the use of length-based Beverton and Holt SSB/R-models is handicapped by underlying hypotheses and sub-models which are restrictive.