Background: Terpenes are industrially relevant natural compounds the biosynthesis of which relies on two well-established-mevalonic acid (MVA) and methyl erythritol phosphate (MEP)-pathways. Both pathways are widely distributed in all domains of life, the former is predominantly found in eukaryotes and archaea and the latter in eubacteria and chloroplasts. These two pathways supply isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), the universal building blocks of terpenes. Results: The potential to establish a semisynthetic third pathway to access these precursors has been investigated in the present work. We have tested the ability of a collection of 93 isopentenyl phosphate kinases (IPK) from the biodiversity to catalyse the double phosphorylation of isopentenol and dimethylallyl alcohol to give, respectively IPP and DMAPP. Five IPKs selected from a preliminary in vitro screening were evaluated in vivo in an engineered chassis E. coli strain producing carotenoids. The recombinant pathway leading to the synthesis of neurosporene and lyco-pene, allows a simple colorimetric assay to test the potential of IPKs for the synthesis of IPP and DMAPP starting from the corresponding alcohols. The best candidate identified was the IPK from Methanococcoides burtonii (UniProt ID: Q12TH9) which improved carotenoid and neurosporene yields ~ 18-fold and > 45-fold, respectively. In our lab scale conditions, titres of neurosporene reached up to 702.1 ± 44.7 µg/g DCW and 966.2 ± 61.6 µg/L. A scale up to 4 L in-batch cultures reached to 604.8 ± 68.3 µg/g DCW and 430.5 ± 48.6 µg/L without any optimisation shown its potential for future applications. Neurosporene was almost the only carotenoid produced under these conditions, reaching ~ 90% of total carotenoids both at lab and batch scales thus offering an easy access to this sophisticated molecule. Conclusion: IPK biodiversity was screened in order to identify IPKs that optimize the final carotenoid content of engineered E. coli cells expressing the lycopene biosynthesis pathway. By simply changing the IPK and without any other metabolic engineering we improved the neurosporene content by more than 45 fold offering a new biosynthetic access to this molecule of upmost importance.