Abstract Purpose The Viking Speech Scale is used to classify speech performance in children with cerebral palsy (CP). A Dutch‐language version (VSS‐NL) has recently become available. This study aimed to determine the reliability and validity of the VSS‐NL and the association with motor type of CP, Gross Motor Function Classification System (GMFCS), Manual Ability Classification System (MACS) and Communication Function Classification System (CFCS). Methods A total of 90 children with CP, recruited throughout the Netherlands, took part in the study. VSS‐NL classifications by speech and language therapists unfamiliar (SLT1) and familiar (SLT2) with the child, parents and physicians were compared. Convergent and discriminant validity were determined with the Intelligibility in Context Scale‐Dutch (ICS‐NL) and the Computer Based Instrument for Low Motor Language Testing (C‐BiLLT). Inter‐ and intrarater reliability were determined by weighted Kappa ( ƙ w ). Validity and associations between VSS‐NL and GMFCS, MACS and CFCS were determined with Spearman's coefficient. Association between VSS‐NL and motor type of CP was determined with Fisher's exact test. Results Interrater reliability was excellent between SLT1‐SLT2 ( ƙ w = 0.93, 95% confidence interval [CI] 0.87–0.99), good between SLT1‐parent ( ƙ w = 0.71, 95% CI 0.60–0.83), SLT1‐physician ( ƙ w = 0.70, 95% CI 0.58–0.81), SLT2‐parent ( ƙ w = 0.71, 95% CI 0.57–0.84), SLT2‐physician ( ƙ w = 0.73, 95% CI 0.62–0.85) and parent‐physician (ƙ = 0.72, 95% CI 0.60–0.85). Intrarater reliability was excellent for SLTs familiar and unfamiliar to the child ( ƙ w = 1.00, 95% CI 1.00–1.00), and very good for physicians ( ƙ w = 0.89, 95% CI 0.75–1.00) and parents ( ƙ w = 0.72, 95% CI 0.62–1.00). Convergent validity was very strong ( r = −0.81, p < 0.001) and discriminant validity moderate ( r = −0.56, p < 0.001). Association with motor type of CP was significant (χ 2 = 27.558, p < 0.001) and strong with GMFCS ( r = 0.62, p < 0.001), MACS ( r = 0.63, p < 0.01) and CFCS ...