Purpose: Speech-language pathologists usually apply a "one size fits all" approach to eliminate compensatory cleft speech characteristics (CSCs). It is necessary to investigate what intervention works best for a particular patient. This pilot study compared the effectiveness of two therapy approaches (a motorphonetic approach and a linguistic-phonological approach) on different subtypes of compensatory CSCs in Dutch-speaking children with a cleft (lip and) palate (CP ± L). Method: Fourteen children with a CP ± L (Mage = 7.71 years) were divided into two groups using block randomization stratified by age, gender, and type of compensatory CSC. Six children received intervention to eliminate anterior oral CSCs (n = 3 motor-phonetic intervention, n = 3 linguistic-phonological intervention). Eight children received intervention to eliminate non-oral CSCs (n = 4 motorphonetic intervention, n = 4 linguistic-phonological intervention). Each child received 10 hr of speech intervention divided over 2 weeks. Perceptual and psychosocial outcome measures were used to determine intervention effects. Results: Children who received linguistic-phonological intervention to eliminate anterior oral CSCs had significantly higher correctly produced consonant scores and health-related quality of life (HRQoL) scores compared to children who received motor-phonetic intervention to eliminate anterior oral CSCs. In the group of children who received intervention to eliminate non-oral CSCs, no significant differences were found in the correctly produced consonant scores nor in the HRQoL scores between the two intervention approaches. Conclusions: Linguistic-phonological intervention seems to be more appropriate to eliminate anterior oral CSCs. The beneficial effects of linguistic-phonological intervention were less pronounced in children with non-oral CSCs. Perhaps, children with non-oral CSCs benefit more from a hybrid phonetic-phonological approach. This study is a step forward in the provision of performance-specific intervention in children with a CP ± L. Replication in larger samples is needed and will aid to tailor treatment plans to the needs of our patients.