The Point Load Test (PLT) was developed as an index test to predict the Uniaxial Compressive Strength (UCS) of core too broken for UCS testing. A correlation factor (K) between UCS and PLT strength of 24 has been proposed although this value is disputed. This paper investigates the value of K for chalk and other weak rocks using published and unpublished data sets and concludes that K is not a unique value even for a single set of specimens but is strength dependent, generally being between 10 and 20 for weaker rock (>25 MPa). It is often less than 10 for rocks of 5 MPa strength or less. Operator error is evaluated and considered to contribute to much of the reported scatter in Point Load Test results. The PLT is considered an appropriate and useful test for predicting the UCS of weak rocks provided that the ISRM methods are strictly observed and that samples with a range of strengths are tested to determine the variation in the correlation factor K with change in strength.