This systematic review compared clinical, service and cost effectiveness of telephone consultations (TC) to video consultations (VC). We searched Embase, CINAHL and MEDLINE for empirical studies that compared TC to VC using clinical, service or economic outcome measures. Clinician or patient preference and satisfaction studies were excluded. Findings were synthesised descriptively. A total of 79 articles were included. The most effective modality was found to be VC in 40 studies (50%) and TC in 3 (4%). VC and TC were found to be equivalent in 28 of the included articles (35%). VC were superior or equivalent to TC for all clinical outcomes. When compared to TC, VC were likely to have better patient engagement and retention, to improve transfer decisions, and reduce downstream sub-acute care utilisation. The impact of telehealth modality on consultation time, completion rates, failure-to-attend rates and acute care utilisation was mixed. VC were consistently found to be more cost effective despite having a higher incremental cost than TC. Our systematic review demonstrates equal or better, but not inferior clinical and cost outcomes for consultations delivered by VC when compared to TC. VC appear to be more clinically effective when visual information is required, when verbal communication with the patient is impaired and when patient engagement and retention is linked to clinical outcomes. We have provided conditions where VC should be used in preference to TC. These can be used by clinicians to guide the choice of telehealth modality. Cost effectiveness is also important to consider when choosing modality.