The variability in the strength of sprayed concrete was investigated and compared to that of counterpart lab-cast concrete. Statistical parameters were utilised in order to analyse the results based on a statistical approach. Two types of sprayed concrete were investigated; plain and fibre reinforced. The variations in the results were examined within the same typeof sprayed concrete and between the two types. The statistical analysis indicated that the strength of the placed sprayed concrete had larger variations compared to lab-cast concrete. The number of replications of test specimens that is required to ensure an acceptable error at certain level of confidence was calculated for various error values and levels of confidence. It was found that while two cubes of lab-cast concrete could be enough to keep the error below 10% at 95% confidence level, 16 cores of sprayed concrete would be necessary. If only three specimens of sprayed concrete were tested, then the expected error could be as high as 25% and 20% at 95% and 90% confidence levels, respectively.