How Many Types of Dinosaurs Were There?

The Number of Dinosaur Species

How many dinosaurs have been discovered and scientifically validated to date?
As of 2025, number of valid non-avian dinosaur genera exceeds 900 to 1,000, with over 1,100 species described. But this figure is not static and is constantly changing with the progress of research.

The number of dinosaur species fluctuates mainly due to the following factors:

1. New dinosaurs continue to be discovered. Currently, new species are being named at a rate of about 30 to 40 per year, and the number of described dinosaurs continues to expand.
2. Through the process of scientific verification, dinosaur genera and species are constantly being organized and revised. Not all dinosaurs named in the past are still considered valid today. In paleontology, there are many "nomina dubia" (doubtful names) that were named based on insufficient fossils and later deemed unidentifiable, as well as "synonyms" that were integrated after being identified as the same species as another dinosaur. A 1990 report by renowned paleontologist Peter Dodson concluded that of the 540 genera described up to that point, only 285 were considered valid. More recent analysis suggests that over 50% of named species may eventually become invalid.

According to the Fukui Prefectural Dinosaur Museum website, "over 1,400 species of dinosaurs have been named, but many did not follow international naming rules or meet the conditions for recognition as a new species, and about 600 of them have become invalid names" (as of October 2025).

Statistical calculations suggest that there were about 1,000, 1,500 genus, or even 2,000 species of dinosaurs.
Not all dinosaurs that existed became fossils. Not all fossils are discovered. The number of discovered dinosaurs is not the number that existed.

There would have been morphological differences due to sex (male and female), as well as physical characteristics that changed with growth within the same species. It is difficult to identify these conditions for non-existent creatures from the remaining fossils.
As research progresses, it may unfortunately be impossible to know exactly "how many species of dinosaurs there were."

Nevertheless, attempts to statistically estimate the number of dinosaurs continue.

Statistical Approaches to Estimating Dinosaur Numbers

Attempts are being made to quantitatively estimate the total number of dinosaurs.

Early Attempts: Dodson's Research (1990)

A 1990 paper by paleontologist Peter Dodson, "Counting dinosaurs: how many kinds were there?", attempted to estimate the number of unknown dinosaurs not just by counting discoveries, but by considering factors like the pace of discovery and the lifespan of species.
This was the application of a "logistic model" to predict future discoveries from the trend of past discovery rates. He estimated the average duration of a single dinosaur "genus" to be about 7.7 million years and used this to calculate how many genera would have existed throughout the Mesozoic Era.

As a result of this analysis, Dodson concluded that the total number of dinosaur genera that existed throughout the Mesozoic would fall in the range of 900 to 1,200.
Since the number of genera he considered valid at the time was 285, this estimate suggested that we only know about 25% of all dinosaurs. This was a landmark study that first quantitatively showed that the majority of dinosaur diversity has yet to be discovered.

The ACE Model: Estimating "Discoverable" Diversity (2006)

A study published by Steve C. Wang and Dodson in the Proceedings of the National Academy of Sciences (PNAS) used a method called the Abundance-based Coverage Estimator (ACE).

The basic idea of the ACE model is to "infer the number of undiscovered items from the frequency of rare discoveries." If excavating new fossils consistently yields only known species, it can be said that the biota of that region is fairly well understood. Conversely, if each excavation reveals rare species found only once, it is likely that many unknown species are still hidden. The ACE model uses data such as the number of "taxa discovered only once" or "taxa discovered only twice" to estimate the number of taxa with a zero discovery probability (i.e., undiscovered). This model estimates the total number of "discoverable" dinosaurs, excluding those that are impossible to discover in principle (left no fossils at all).

The TRiPS Model: A Probabilistic Approach to Species-Level Richness (2016)

Jostein Starrfelt and Lee Hsiang Liow introduced a new statistical model, True Richness estimated using a Poisson Sampling (TRiPS).

Whereas ACE focuses on the discovery frequency (abundance) of taxa, TRiPS attempts to estimate the inherent "sampling probability" of each species based on how many times it has been observed in the fossil record. For example, a species known from many specimens is considered to have a high sampling probability because it was either originally abundant or had characteristics that made it more likely to be preserved as a fossil. This model corrects for sampling bias by considering these differences in sampling probabilities among species to estimate diversity.

A notable feature of the TRiPS model is that it directly estimates the number at the "species" level, rather than the genus level. Applying this model to the dinosaur fossil record throughout the Mesozoic Era resulted in an estimated total of 1,936 species (with a 95% confidence interval of 1,543–2,468 species).

Furthermore, this study delved into the diversity of major dinosaur groups. The results concluded that Theropoda, which led to birds, was the most species-rich group (estimated at about 1,115 species), followed by Sauropodomorpha (estimated at about 513 species) and Ornithischia (estimated at about 508 species).